INDEX

Explanations

quotations in the format of a person speaking

oai_token-act-pair · gpt-3.5-turbo

phrases indicating uncertainty or quotes from individuals

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 globally

-0.77

ä»£

-0.76

 ç¥ŀ

-0.76

ãĥİ

-0.76

eaturing

-0.73

 optimized

-0.72

 worldwide

-0.72

 nutshell

-0.71

 domestically

-0.71

levant

-0.71

POSITIVE LOGITS

 Paddock

1.10

 Jaune

1.05

 Manziel

1.03

 Wenger

1.02

 Korra

1.01

 Pyrrha

1.01

 Elsa

0.99

 Klopp

0.99

 Cosby

0.97

Sgt

0.96

Activations Density 0.466%

quotations in the format of a person speaking

phrases indicating uncertainty or quotes from individuals

No Comments

No Known Activations